Root/
| 1 | - 4.5 V rail was intended to provide the ADXL32x with a supply voltage |
| 2 | that's at a safe distance from its 2.4 V minimum. However, the current |
| 3 | circuit wouldn't work with the acceleration sensor operating at 4.5 V. |
| 4 | It turns out the system operates well enough with the 4.5 V rail joined |
| 5 | to the 3 V rail. |
| 6 | |
| 7 | - the 6 V rail drops to 5 V after relatively little use. Need to improve |
| 8 | the power supply. |
| 9 | |
| 10 | - connections between boards are sensitive to mechanical forces and break |
| 11 | easily in the high-g environment we have here. Should consider moving |
| 12 | to an FPC at least for the LEDs. A stiff support beam (e.g., an |
| 13 | aluminium rail) helps to reduce mechanical stress. |
| 14 | |
| 15 | - the vertical resolution is a bit coarse. We could improve it and make |
| 16 | the whole system simpler by adding local demultiplexing with 74HC594 |
| 17 | or 75HC595. Such chips could drive the LEDs directly. |
| 18 | |
| 19 | - the battery pack needs room for a plug. (The original design assumed |
| 20 | the battery pack would be mounted on the stick, but this now seems |
| 21 | undesirable, due to the large weight. Current stick weight is about |
| 22 | 120 g and each AA battery would add about 25 g.) |
| 23 | |
| 24 | - the system really ought to have a button or even a joystick for user |
| 25 | input. |
| 26 | |
| 27 | - once a button is added, we could have local-only operation with images |
| 28 | coming from Flash. |
| 29 |
Branches:
master
tornado-v1
